US Patent Issued to Carnegie Mellon University on April 1 for "System and method for solving missing annotation object detection" (Pennsylvania Inventors)

ALEXANDRIA, Va., April 2 -- United States Patent no. 12,266,156, issued on April 1, was assigned to Carnegie Mellon University (Pittsburgh).

"System and method for solving missing annotation object detection" was invented by Marios Savvides (Wexford, Pa.), Zhiqiang Shen (Pittsburgh), Fangyi Chen (Pittsburgh) and Han Zhang (Pittsburgh).

According to the abstract* released by the U.S. Patent & Trademark Office: "Disclosed herein is a system and method for improving the accuracy of an object detector when trained with a dataset having a significant number of missing annotations. The method uses a novel Background Recalibration Loss (BRL) which adjusts the gradient direction according to its own activation to reduce the adverse effect of erro...